© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 22-TRANSCODER-HP
    7. 97754
    Prev
    Next
    INDEX
    Explanations

    Code and natural language mix

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    thag
    -0.28
    yntax
    -0.27
    æĦıåĽ¾
    -0.25
    å·²ä¹ħçļĦ
    -0.25
    èĤ®
    -0.25
     Drag
    -0.24
    shed
    -0.24
    å¢ĥçķĮ
    -0.24
    unsch
    -0.24
    venta
    -0.23
    POSITIVE LOGITS
    finally
    0.25
    orget
    0.25
    olith
    0.24
     кÑĢÑĥ
    0.24
    /cop
    0.24
    andre
    0.23
    ensus
    0.23
     finally
    0.23
    evin
    0.23
    ICA
    0.22
    Activations Density 0.047%

    No Known Activations